ppt

34
Indranil Gupta Lecture 9 Peer-to-peer systems Computer Science 425 Distributed Systems

description

 

Transcript of ppt

Page 1: ppt

Indranil Gupta

Lecture 9

Peer-to-peer systems

Computer Science 425Distributed Systems

Computer Science 425Distributed Systems

Page 2: ppt

Peer to peer systems(Next few lectures)

D.S. Theory

2 Angles of Distributed Systems

Page 3: ppt

Some Questions

• Why do people get together?– to share information – to share and exchange resources they have

• books, class notes, experiences, music cd’s

• How can computers help people – find information– find resources– exchange and share resources

Page 4: ppt

• Existing technologies: The Web!– Search engines– Forums: chat rooms, blogs, ebay– Online business

• But, the web is heavy weight if you want specific resources: say a Beatles’ song “PennyLane”

• A google search will give you their bio, lyrics, chords, articles on them, and then perhaps the mp3

• But you want only the song, nothing else!• If you can find a peer who wouldn’t mind exchanging her

Beatles songs for your Miles Davis recordings, that would be great! – Napster: a light weight solution (lighter than the Web)

Page 5: ppt
Page 6: ppt

A Brief History• [6/99] Shawn Fanning (freshman Northeastern U.) releases

Napster online music service• [12/99] RIAA sues Napster, asking $100K per download• [3/00] 25% UWisc traffic Napster, many universities ban it• [00] 60M users• [2/01] US Federal Appeals Court: users violating

copyright laws, Napster is abetting this• [9/01] Napster decides to run paid service, pay % to

songwriters and music companies• [Today] Napster protocol is open, people free to develop

opennap clients and servers http://opennap.sourceforge.net

• [Today] eDonkey, BitTorrent, …

Page 7: ppt

Napster Structure

S

S

S

P

P

PP

P

P

Client machines (“Peers”)

napster.com Servers

Store their ownfiles

Store a directory, i.e., filenames with peer pointers

Filename Info about

PennyLane.mp3 Beatles, @ 128.84.92.23:1006

…..

Page 8: ppt

Napster OperationsClient• Connect to a Napster server• Upload list of music files that you want to share (names only, not the

files themselves!)– Server maintains list of

<filename, ip_address, portnum> tuples

• Search from a client:– Send server keywords to search with– (Server searches its directory with the keywords)– Server returns a list of matching hosts - <ip_address, portnum> tuples to client– Client pings each host in the list to find transfer rates – Client fetches file from best host

• All communication uses TCP

Page 9: ppt

Napster Search

S

S

S

P

P

PP

P

P

Peers

napster.com Servers

Store their ownfiles

Store peer pointers for all files

3. Response 1. Query

2. All servers search their lists (ternary tree algo.)

4. ping candidates5. download from best host

Page 10: ppt

Problems

• Centralized server a source of congestion

• Centralized server single point of failure

• No security: plaintext messages and passwds

• napster.com responsible for abetting users’ copyright violation– “Indirect infringement”

Page 11: ppt

Gnutella

• Eliminate the servers• Client machines search and retrieve amongst

themselves• Clients act as servers too, called servents• [3/00] release by AOL, immediately withdrawn,

but 88K users by 3/03• Original design underwent several modifications;

we’ll look at the initial version

http://www.limewire.com

Page 12: ppt

Gnutella

P

P

P

P

PP

Servents (“Peers”)

PConnected in an overlay graph

Store their ownfiles

Also store “peer pointers”

a peer pointer

Page 13: ppt

How do I search for my Beatles file?

• Gnutella routes different messages within the overlay graph

• Gnutella protocol has 5 main message types– Query– QueryHit (response to query)– Ping (to probe network for other peers)– Pong (reply to ping, contains address of another peer)– Push (used to initiate file transfer)

• We’ll go into the message structure and protocol now(note: all fields except IP address are in little-endian format)

Page 14: ppt

Descriptor ID Payload descriptor TTL Hops Payload length

Descriptor Header

Type of payload0x00 Ping0x01 Pong0x40 Push0x80 Query0x81 Queryhit

Decremented at each hop,Message dropped when ttl=0ttl_initial usually 10

Incremented at each hop

ID of this searchtransaction

Number of bytes ofmessage following this header

0 15 16 17 18 22

Payload

Gnutella Message Header Format

Page 15: ppt

Minimum Speed Search criteria (keywords)

Query (0x80)

0 1 …..

Payload Format in Gnutella Query Message

Page 16: ppt

Gnutella Search

P

P

P

P

PP

PWho has PennyLane.mp3?

Query’s flooded out, ttl-restricted, forwarded only once

TTL=2

TTL=1

Page 17: ppt

Num. hits port ip_address speed (fileindex,filename,fsize) servent_id

0 1 3 7 11 n n+16

QueryHit (0x80) : successful result to a query

Results

Unique identifier of responder;a function of its IP address

Info aboutresponder

Payload Format in Gnutella Query Reply Message

Page 18: ppt

Gnutella Search

P

P

P

P

PP

PWho has PennyLane.mp3?

Successful results QueryHit’s routed on reverse path

Page 19: ppt

Avoiding excessive trafficTo avoid duplicate transmissions, each peer maintains a list of

recently received messages• Query forwarded to all neighbors except peer from which

received (and this is remembered)• Each Query (identified by DescriptorID) forwarded only once • QueryHit routed back only to peer from which Query received

with same DescriptorID• Duplicates with same DescriptorID and Payload descriptor

(msg type) are dropped• QueryHit with DescriptorID for which Query not seen is

dropped

Page 20: ppt

After receiving QueryHit messages• Requestor chooses best QueryHit responder

– Initiates HTTP request directly to responder’s ip+port

GET /get/<File Index>/<File Name>/HTTP/1.0\r\n

Connection: Keep-Alive\r\nRange: bytes=0-\filesize\nUser-Agent: Gnutella\r\n\r\n

• Responder then replies with file packets following the msg:HTTP 200 OK\r\n

Server:Gnutella\r\nContent-type:application/binary\r\nContent-length: 1024 \r\n\r\n

• HTTP is the file transfer protocol. Why?• Why the “range” field in the GET request?• What if responder is behind firewall that disallows incoming connections?

Page 21: ppt

Dealing with Firewalls

P

P

P

P

PP

P

Requestor sends Push to responder asking for file transfer

Has PennyLane.mp3But behind firewall

(Why is the Push routed and not sent directly?)

Page 22: ppt

servent_id fileindex ip_address port

Push (0x40)

same as in received QueryHit Address at which

requestor can acceptincoming connections

Page 23: ppt

• Responder establishes a TCP connection at ip_address,port specified. Sends

GIV <File Index>:<Servent Identifier>/<File Name>\n\n

• Requestor then sends GET to responder (as before) and file is transferred

• What if requestor is behind firewall too?– Gnutella gives up– Can you think of an alternative solution?

Page 24: ppt

Ping-Pong

•Peers initiate Ping’s periodically

•Ping’s flooded out like Query’s, Pong’s routed along reverse path like QueryHit’s

•Pong replies used to update set of neighboring peers

Port ip_address Num. files shared Num. KB shared

Pong (0x01)

Ping (0x00)no payload

Page 25: ppt

Gnutella Summary• No servers • Peers/servents maintain “neighbors”, this forms an overlay graph• Peers store their own files• Queries flooded out, ttl restricted• Query Replies reverse path routed• Supports file transfer through firewalls (one-way)• Periodic Ping-pong to keep neighbor lists fresh in spite of peers

joining, leaving and failing– List size specified by human user at peer : heterogeneity means some

peers may have more neighbors– Gnutella found to follow power law distribution:P(#neighboring links for a node = L) ~ (k constant)

kL

Page 26: ppt

Problems• Ping/Pong constitutes 50% traffic

– Solution: Multiplex, cache and reduce frequency of pings/pongs

• Repeated searches with same keywords– Solution: Cache Query and QueryHit messages

• Modem-connected hosts do not have enough bandwidth for passing Gnutella traffic– Solution: use a central server to act as proxy for such

peers– Another solution:

FastTrack System (in a few slides)

Page 27: ppt

Problems (contd.)

• Large number of freeloaders– 70% of users in 2000 were freeloaders

• Flooding causes excessive traffic– Is there some way of maintaining meta-information

about peers that leads to more intelligent routing?

Structured Peer-to-peer systems

e.g., Chord System (next lecture)

Page 28: ppt

FastTrack

• Hybrid between Gnutella and Napster• Takes advantage of “healthier” participants in the

system (higher bandwidth nodes, nodes that are around most of the time, nodes that are not freeloaders, etc.)

• Underlying technology in Kazaa, KazaaLite, Grokster

• Proprietary protocol, but some details available• Like Gnutella, but with some peers designated as

supernodes

Page 29: ppt

A FastTrack-like System

PP

P

PPeers

S

S

SupernodesP

Page 30: ppt

FastTrack (contd.)• A supernode stores a directory listing

(<filename,peer pointer>), similar to Napster servers• Supernode membership changes over time• Any peer can become (and stay) a supernode,

provided it has earned enough reputation– Kazaalite: participation level of a user between 0 and

1000, initially 10, then affected by length of periods of connectivity and total number of uploads from that client

• A peer searches for a file by contacting a nearby supernode

Page 31: ppt

Wrap-up NotesApplies to all p2p systems• How does a peer join the system (bootstrap)

– Send an http request to well known url http://www.myp2pservice.com

– Message routed after DNS lookup to a well known server that has partial list of recently joined peers, and uses this to initialize new peers’ neighbor table

• Lookups can be speeded up by having each peer cache:– Queries and their results that it sees– All directory entries (filename, host) mappings that it sees– The files themselves

Page 32: ppt

Summary• Napster: protocol overview, more details available on

webpage• Gnutella protocol• FastTrack protocol• Protocols continually evolving, software for new clients

and servers conforming to respective protocols: developer forums at – Napster: http://opennap.sourceforge.net– Gnutella: http://www.limewire.com

• Others– Peer to peer working groups: http://www.p2pwg.com

Page 33: ppt

DHT=Distributed Hash Table• A hash table allows you to insert, lookup and

delete objects with keys• A distributed hash table allows you to do the same

in a distributed setting (objects=files)• Napster, Gnutella, FastTrack are all DHTs (sort

of)• So is Chord, a structured peer to peer system that

we study next lecture

Page 34: ppt

• HW 2 due today

• HW3 out today, due Sep 27 (watch newsgroup and website)

• Midterm on Oct 4 (Thursday), will include all topics covered in HW1-HW3